Intuit launches lab for employees to experiment with generative AI, will share lessons at VB Transform

Financial services company Intuit has long been at the forefront of integrating AI into its suite of products, including TurboTax and QuickBooks. Now the company is poised to harness the power of generative AI, the groundbreaking technology that’s rapidly transforming the enterprise world.

This month, the company launched “GenStudio,” a generative AI lab for employees from various backgrounds – engineers, data scientists and product managers – to explore and experiment with generative AI on a larger scale. This initiative seeks to harness the creative potential of Intuit’s workforce, and aims to develop GenAI in a systematic manner, incorporating essential safeguards against the risks of generative AI within its technology architecture from the outset.

Ashok Srivastava, Intuit’s Chief Data Officer, is set to speak at the VB Transform event in San Francisco on July 11 and 12. He will share insights into the company’s journey to tap into the potential of generative AI and discuss the lessons learned from establishing GenStudio. Generative AI gained widespread attention last year with the release of OpenAI’s ChatGPT, a powerful chatbot that mimics human conversation. Intuit’s participation at Transform  highlights the event’s commitment to showcasing the latest enterprise innovations in AI and generative AI implementation.

In recent interviews with Srivastava and Intuit’s Chief Technology Officer, Mariana Tessel, the executives revealed their strategies for staying at the forefront of generative AI in 2023 and beyond. Their insights provide a glimpse into the story that Intuit will share at the Transform event and offer a blueprint for other companies looking to stay on the cutting edge of technology. My interview with Mariana is embedded here:

Last year, Tessel wowed attendees at the Transform conference when she revealed that Intuit had implemented two million AI models, which are refreshed daily. This large-scale deployment enables the company to offer personalized AI experiences for its millions of customers.

Intuit has been utilizing large language models (LLMs), the technology underpinning ChatGPT, for several years, specifically in the area of tax information. The company experimented with an early version of OpenAI’s GPT technology (known as GPT-3) when it was first released via an API in 2021. According to Srivastava, this technology reduced customer call time by one million hours per year, thanks to its ability to summarize calls for both experts and customers.

​​Confident in the technology’s potential, Intuit established the generative AI lab, GenStudio, to help internal employees scale generative AI while incorporating safeguards against potential risks. This lab represents a natural progression from the corporate machine learning platforms that emerged during the initial wave of AI adoption, such as Uber’s Michelangelo, LinkedIn’s Pro-ML, and eBay’s Krylov.

While Srivastava remains tight-lipped about specific results ahead of his Transform presentation, he did share a tantalizing teaser: “We’ve done some preliminary testing of additional generative AI capabilities that are giving extraordinary results. I almost couldn’t believe it when I saw how well they were working.”

Here are some other preliminary conclusions Intuit has reached, which its executives say are relevant for peers seeking to embrace generative AI.

Intuit is confident in maintaining its leadership in its domain of expertise, despite potential competition from ChatGPT. Although tax codes are readily available for AI engines to analyze and disseminate, for example, Intuit’s years of building a robust knowledge base ensure accuracy in its LLM outputs.

Another lesson is the importance of having a forward-looking internal team and a “platform” for experimenting with emerging technology trends. This approach allows companies to quickly adapt to new advancements like generative AI, providing a significant advantage over competitors.

The significance of a unified data strategy for internal stakeholders, which ensures a single source of truth across the company. For Intuit, this includes the “Data Exchange,” which processes data from over 24,000 financial institutions.